Aiming to use the Halifax effectively as a daytime bomber, the Air Staff prompted Handley Page to create a high-altitude variant equipped with either Bristol Hercules VIII or Rolls-Royce Merlin 60 engines, both featuring two-stage superchargers capable of maximum power at 30,000 feet. These engines were urgently developed for the Wellington V and VI with pressurized cabins. However, the Hercules VIII experienced unsolvable supercharger surging issues and was discontinued. The Merlin 60, 61, and 62 engines entered production for Spitfires and Mosquitos and were proposed for the Halifax IV, which was to have a reinforced bomb floor, larger bomb doors to house 4,000-lb and 8,000-lb bombs, extended wingtips increasing the span to 104 feet, and enlarged symmetrical (E-type) fins and rudders. Handley Page prioritized the Merlin 60 over the Hercules, designating HR756 as the Halifax IV prototype. Although built with Merlin 60 engines, only Merlin XXs were available for the initial flight tests in March 1943, resulting in it being labeled as Halifax II series 2. It was planned to upgrade to Merlin 60s when available, but the aircraft was used for engine development at Rolls-Royce Ltd in Hucknall, retaining its original 99-ft wingspan despite eventual installations of Merlin 61 and 65 engines at different times.
On August 5, 1941, Handley Page recommended to Air Marshal F. J. Linnell the early production of some Halifax IV features. However, in September, the Controller of Research and Development requested the integration of Bristol Hercules VI or XVI engines into current production aircraft as soon as possible. With structural adjustments to increase the maximum weight to 64,000 lb, this version, the Halifax III (H.P.61), was estimated to reach 307 mph at 21,000 feet. Improved Hercules 100 engines were hoped to replace the failed Hercules VIII. Handley Page expressed concerns about numerous modifications interfering with production unless delayed until the 201st Halifax. These included diverse features like target towing, paratrooping, heavy-gauge protective engine cowling, multiple flare chutes, revised ammunition tracks to the tail turret to clear the paratroop exit cone, a revised dual-control pilot’s seat, rocket gear for short take-off, and arrestor gear for short landing. He doubted these modifications would enhance the Halifax’s efficiency as a bomber, although he supported fitting a B-P Type A mid-upper turret and F.N.64 ventral turret for daylight operations. Linnell responded that he could not guarantee no further modifications until the 200th aircraft but could defer all except exhaust flame dampers, engine rear armor, astrograph and D/F loop, fixed fittings for 4,000-lb and 8,000-lb bombs, increased electrical generation, and glider towing. Handley Page accepted these and offered Rawcliffe aerodrome for the initial towed flight trials of the new Slingsby Hengist paratroop glider, though these trials were ultimately conducted by the Airborne Forces Experimental Establishment at Sherburn-in-Elmet.
In January 1942, R9534, just completed as a prototype Mk.II series 1 (Special) with a Tempsford nose and B-P Type C mid-upper turret, was chosen for the trial installation of Hercules VI engines, making it the first prototype Halifax III, flown on October 12, 1942, with daylight camouflage above and ‘trainer yellow’ below. In December, the B-P Type C turret was replaced with a Defiant-type B-P A Mk.III. Further performance improvements were achieved in February 1943 by retracting the tailwheel and fitting Beaufighter-type Hercules cowlings in preparation for Cricklewood’s 700th production Halifax, HX226, the first true Mark III, flown by Talbot in August 1943. More Halifax IIIs were produced than any other variant, totaling 2,127, with 2,091 delivered before VJ-Day. These included refinements like a return to a retractable tailwheel after overcoming runway shimmy issues. Halifax IIIs were first issued in November 1943 to No.433 Squadron, RCAF, and No.466 Squadron, RAAF, eventually equipping forty-one operational squadrons and numerous training and experimental units.